| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 19.99591836734694, | |
| "global_step": 2440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.4127440452575684, | |
| "eval_runtime": 162.9527, | |
| "eval_samples_per_second": 6.64, | |
| "eval_steps_per_second": 0.835, | |
| "eval_wer": 1.0, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 2.9441540241241455, | |
| "eval_runtime": 163.1192, | |
| "eval_samples_per_second": 6.633, | |
| "eval_steps_per_second": 0.834, | |
| "eval_wer": 1.0, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.3244493212689078, | |
| "eval_loss": 1.2275432348251343, | |
| "eval_runtime": 159.0312, | |
| "eval_samples_per_second": 6.804, | |
| "eval_steps_per_second": 0.855, | |
| "eval_wer": 0.8635921622015851, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 0.0002388, | |
| "loss": 3.9307, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.2023591638341556, | |
| "eval_loss": 0.7168374061584473, | |
| "eval_runtime": 158.5869, | |
| "eval_samples_per_second": 6.823, | |
| "eval_steps_per_second": 0.858, | |
| "eval_wer": 0.501920552341129, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.17312679219727137, | |
| "eval_loss": 0.617878794670105, | |
| "eval_runtime": 158.2056, | |
| "eval_samples_per_second": 6.839, | |
| "eval_steps_per_second": 0.86, | |
| "eval_wer": 0.41668692565760684, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.16569868368602203, | |
| "eval_loss": 0.6016563177108765, | |
| "eval_runtime": 157.9983, | |
| "eval_samples_per_second": 6.848, | |
| "eval_steps_per_second": 0.861, | |
| "eval_wer": 0.37708465016774445, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "learning_rate": 0.00025391752577319586, | |
| "loss": 0.6036, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.16388987711259606, | |
| "eval_loss": 0.6307795643806458, | |
| "eval_runtime": 159.8432, | |
| "eval_samples_per_second": 6.769, | |
| "eval_steps_per_second": 0.851, | |
| "eval_wer": 0.3708610881509214, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.1604865775407946, | |
| "eval_loss": 0.6786409616470337, | |
| "eval_runtime": 160.9659, | |
| "eval_samples_per_second": 6.722, | |
| "eval_steps_per_second": 0.845, | |
| "eval_wer": 0.35467010259153015, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.15351281380865225, | |
| "eval_loss": 0.696066677570343, | |
| "eval_runtime": 165.3737, | |
| "eval_samples_per_second": 6.543, | |
| "eval_steps_per_second": 0.822, | |
| "eval_wer": 0.3348324986629066, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.83, | |
| "learning_rate": 0.00019206185567010307, | |
| "loss": 0.303, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.15007951033634373, | |
| "eval_loss": 0.713607132434845, | |
| "eval_runtime": 160.5614, | |
| "eval_samples_per_second": 6.739, | |
| "eval_steps_per_second": 0.847, | |
| "eval_wer": 0.3272961540331599, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.1534270883786321, | |
| "eval_loss": 0.724854052066803, | |
| "eval_runtime": 158.4702, | |
| "eval_samples_per_second": 6.828, | |
| "eval_steps_per_second": 0.858, | |
| "eval_wer": 0.3330821218456751, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.15021667102437603, | |
| "eval_loss": 0.7961975932121277, | |
| "eval_runtime": 158.6964, | |
| "eval_samples_per_second": 6.818, | |
| "eval_steps_per_second": 0.857, | |
| "eval_wer": 0.3294598142655711, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.15035811798390933, | |
| "eval_loss": 0.8382342457771301, | |
| "eval_runtime": 159.3031, | |
| "eval_samples_per_second": 6.792, | |
| "eval_steps_per_second": 0.854, | |
| "eval_wer": 0.3209996596489522, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.11, | |
| "learning_rate": 0.0001302061855670103, | |
| "loss": 0.1685, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.1506152942739699, | |
| "eval_loss": 0.8463586568832397, | |
| "eval_runtime": 159.0294, | |
| "eval_samples_per_second": 6.804, | |
| "eval_steps_per_second": 0.855, | |
| "eval_wer": 0.3242329946030048, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.14896507974608128, | |
| "eval_loss": 0.9096932411193848, | |
| "eval_runtime": 162.4549, | |
| "eval_samples_per_second": 6.66, | |
| "eval_steps_per_second": 0.837, | |
| "eval_wer": 0.3218019156901833, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.14725485741717853, | |
| "eval_loss": 0.963500440120697, | |
| "eval_runtime": 158.9025, | |
| "eval_samples_per_second": 6.809, | |
| "eval_steps_per_second": 0.856, | |
| "eval_wer": 0.31263674818884624, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.39, | |
| "learning_rate": 6.835051546391752e-05, | |
| "loss": 0.1031, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.1464233207459827, | |
| "eval_loss": 0.9706696271896362, | |
| "eval_runtime": 158.7074, | |
| "eval_samples_per_second": 6.818, | |
| "eval_steps_per_second": 0.857, | |
| "eval_wer": 0.3110565468955122, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.14608041902590194, | |
| "eval_loss": 1.0005509853363037, | |
| "eval_runtime": 158.0719, | |
| "eval_samples_per_second": 6.845, | |
| "eval_steps_per_second": 0.86, | |
| "eval_wer": 0.31331745028443625, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.14561750170379292, | |
| "eval_loss": 1.0291253328323364, | |
| "eval_runtime": 159.3838, | |
| "eval_samples_per_second": 6.789, | |
| "eval_steps_per_second": 0.853, | |
| "eval_wer": 0.31030291243253755, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.67, | |
| "learning_rate": 6.494845360824742e-06, | |
| "loss": 0.0629, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 0.1464233207459827, | |
| "eval_loss": 1.0326457023620605, | |
| "eval_runtime": 158.8998, | |
| "eval_samples_per_second": 6.809, | |
| "eval_steps_per_second": 0.856, | |
| "eval_wer": 0.31287985608012836, | |
| "step": 2440 | |
| } | |
| ], | |
| "max_steps": 2440, | |
| "num_train_epochs": 20, | |
| "total_flos": 7.698957470096574e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |